The two patients we present had radiographically occult disease. This is similar to what was reported by Fan et al [2]. These data high light the need for through investigations of patients with breast cancer with new symptoms. An endocrine approach was used for both patients since there was no imminent severe end organ failure. The tumor flare reaction could have been misinterpreted as disease progression. It is critical to be aware of this phenomenon and to avoid premature discontinuation of a valuable treatment. Tumor flare was recognized as an event during successful treatment of advanced breast cancer over 45 years ago [12]. Flare is a transient increase in symptoms and signs of hormonally sensitive breast cancer. It may be accompanied by severe hypercalcemia and other electrolyte disturbances. Supportive care and reassurance can manage most patients’ issues. Letrozole has been reported to cause a similar event [13]. Aromatase inhibitors with Palbociclib have as been noted to cause a pseudoprogression of breast cancer. This is the first report of fulvestrant and everolimus causing a tumor flare reaction.

     Sun et al [1] reported 19 cases of bone marrow metastases from breast cancer seen in their cancer center. They noted all cases were ER positive, the incidence of skeletal adverse events was higher in this group and survival was shorter in this group compared to patients with bone metastases. Pain was an adverse finding. Combination therapy in bone marrow metastases was numerically superior to single agent treatment, 21 vs 5 months. Fan et al [2] reported a case of diffuse bone marrow metastases as the initial clinical presentation of occult estrogen positive breast cancer. The patient’s initial complaint was headache. Atwal et al [3] also reported a case of estrogen receptor positive diffuse bone marrow metastases occurring 6 years after the initial diagnosis. Kopp et al [4] reported 22 cases of breast cancer patients with bone marrow metastases. These 22 cases represented 0.17% of all the breast cancer patients seen over a 14-year time span. All patients had other sites of metastases. Only 4 of the 22 were negative for hormone receptors. Over half were invasive ductal carcinoma. Five patients had abnormal peripheral blood smears. Both patients we report had areas of bone pain, estrogen and progesterone receptor positivity and neither had abnormal blood smears. One patient presented with anemia and the other developed it after treatment was started. The patient with less extensive disease had anemia. This is consistent with the observations of Kopp et al [4] that the blood smear abnormalities did not correlate with the extent of disease.

     Bitter et al [5] noted extreme difficulty with the accurate diagnosis of bone marrow metastases by lobular breast cancer. Reasons cited for the poor specificity and sensitivity of diagnosis by hematoxylin-eosin-stained bone marrow samples included 4 issues. First off lobular carcinoma cells are similar in size to hematopoietic cells. Lobular carcinoma infiltrations often include single cells, the cells lack distinctive features and there is limit normal tissue reaction to the presence of the cancer cells. They recommend pancytokeratin stains for all bone marrow samples from breast cancer patients. Clarke and Cheung [6] reported a case of lobular carcinoma metastatic to the bone marrow. Their illustrations of the peripheral smear and marrow emphasize the difficulties in diagnosis reported by Bitter et al [5]. Anecdotally a patient with a similar presentation of lobular carcinoma had over 1500 cells/ml by circulating tumor cell assay but the peripheral smears were repeatedly read as normal (personal observations). Neither one of the patients we are reporting had abnormal peripheral blood smears collected on multiple occasions.These peripheral blood smears were reviewed by multiple Pathologists.

     He et al [7] reviewed the unusual sites of metastases seen in patients with lobular carcinoma. They noted diffuse marrow replacement by lobular carcinoma may cause signal reversal of bone marrow on T-1 weighted MRI. On FDG-PET scans diffuse marrow metastases may appear as numerous “hot spots” throughout the skeleton. Neither one of the patients we are reporting had CT/PET, bone scan or MRI of skull findings pointing to bone marrow metastases.

     Akagi et al [8] reported the successful control of estrogen positive lobular breast carcinoma with metastasis to multiple sites including bone marrow using Adriamycin and Cytoxan. The patient’s cytopenias and disease burden were well controlled. She progressed 15 months later and received erbulin with good results. Pahouja el al [9] also noted success with Adriamycin in a very similar patient. The 2 patients in this report only had bone marrow metastasis. Certainly, bone marrow metastasis can be viewed as a visceral crisis and treated with cytotoxic chemotherapy but endocrine treatment was successful in the 2 patients reported herein.

     Nakagawa et al [10] reported the control of cytopenias from widely metastatic lobular carcinoma of the breast involving the bone marrow initially with letrozole. When the patient progressed, control was achieved with fulvestrant and Palbociclib. Garufi et al [11] reported a similar case in which control of the patient’s disease was achieved with leuprolein, letrozole and Palbociclib. The 2 patients in this report both developed metastasis while on adjuvant anastrazole. Initial responses to fulvestrant and abemaciclib were salutary but both progressed on therapy. Both have improved on using everolimus in place of the CDK 4/6 inhibitor. Gao et al [15] noted the responses of metastatic lobular breast carcinoma to combined endocrine and CDK 4/6 inhibitor treatment 6.9 month improvement in progression free survival over single agent endocrine therapy. Progression free survival improvements in other breast cancer subgroups were numerically superior at 9.1-9.6 months. The responses of our 2 patients to CDK 4/6 inhibition were also relatively short.

     Martin et al [16] reported data on second line treatment in hormone receptor positive breast cancer patients failing endocrine therapy and a CDK4/6 inhibitor. Chemotherapy was the most frequent choice. Treatment with everolimus was associated with an improved overall survival compared to chemotherapy while single agent fulvestrant was not. Lee et al [17] looked-for genomic signals of resistance to CDK 4/6 inhibitor treatment in hormone receptor positive breast cancer. They noted FGFR1 amplification, PTEN loss and DNA repair pathway gene mutations showed significant associations with shorter progression free survival for patients receiving CDK 4/6 inhibitor therapy. Some of these mutations activate downstream MAPK/AKT/Mtor which confers resistance to CD4K 4/6 inhibition. These observations hint at the mechanism the positive role for everolimus in combination with fulvestrant that we report for our 2 patients.